Solenoid Players
By Robbie Rhodes
Laurent, I hope that you have a chance to see a Boesendorfer SE piano, and to decide for yourself about the capability of this technology.
The SE mechanism was designed to faithfully reproduce a performance recorded on _any_ piano equipped with the same system. A self- calibration procedure assures that the hammers in all pianos respond identically at all possible velocities, and a companion manual procedure is performed to verify that all of the pedal actions are identical amongst SE instruments. The positions of the keyframe shifter and the damper-lifting tray are recorded and reproduced using linear position transducers.
Hammer velocity is measured by precision shutters passing through the light beams of opto-interrupters. The solenoid velocity is measured with a permanent-magnet linear transducer. Hammer-velocity commands are presented to the "key motor", which consists of a large solenoid and its velocity transducer; the current to the solenoid is adjusted by a feedback controller to yield the commanded velocity.
Since the solenoid (hence the key) is velocity-controlled, the varying friction inherent in a piano action has little effect on the player mechanism. It is possible to depress every key reliably without any sound produced, by issuing the command to strike the key with zero velocity. The SE mechanism indeed performs this action: the key goes down, the damper rises, and the hammer swings up to the let-off position without striking the string. All 88 keys (97 keys in the *big* Boesen- dorfer) will successfully pass this test, every time, after calibration.
The performance capabilities of the SE mechanism are tremendous; so is its cost. The solenoid action stack is so heavy that an automotive transmission jack must be used to safely install it ! The system was designed in the mid-1970s, yet after twenty years it still represents the ultimate standard of performance against which all other reproducing systems may be compared.
Best regards, Robbie Rhodes
|
(Message sent Wed 5 Mar 1997, 02:17:49 GMT, from time zone GMT-0800.) |
|
|